24th ETSF Workshop on Electronic Excitations: Light-Matter Interaction and Optical Spectroscopy from Infrared to X-Rays
16 – 20 September 2019
Organizers: Claudia Rödl, Gabriele D’Avino, Michiel van Setten, Guido Fratesi, Elena Cannuccia, Carina Faber
Funding and support: Friedrich-Schiller-Universität Jena, Psi-k, Michael Stifel Center Jena, Abbe Center of Photonics Jena, SFB NOA, jenaparadies.de
The 24th ETSF Workshop on Electronic Excitations focused on light-matter interaction and the theoretical description of spectroscopic techniques that nowadays probe electrons, plasmons, excitons, and phonons across different energy and time scales with unprecedented accuracy. A deep physical understanding of the underlying quantum many-body effects is of paramount importance to analyze these experimental observations, extract the wealth of information therein, and ultimately render theoretical simulations predictive.
The workshop aimed at discussing the most recent advances in the theoretical description of the interaction between light and matter focusing on first-principles methods. This broad subject was covered in its diverse declinations, from core-level spectroscopy to collective low-energy excitations, discussing also matter under extreme conditions, and systems driven out of equilibrium by strong laser pulses. The workshop has fostered a fruitful exchange between theorists and experimentalists, opening new horizons towards the next generation of novel spectroscopy techniques. The workshop also tackled the challenges posed by the formidable complexity of heterogeneous and nano-structured systems such as those of interest for light harvesting and energy generation, prompting to bridge the gap between experimental and in silico spectroscopy.
The main topics were:
- Linear and non-linear optical spectroscopy
- Core-level spectroscopies
- Ultrafast excitation dynamics
- Electron-phonon coupling
- Light harvesting in natural and synthetic systems
The workshop took place in the Rosensäle conference building of the Friedrich-Schiller-Universtität Jena in the historical center of the town of Jena (Germany). It was attended by 102 scientists, thereof many young researchers. The organizers are very pleased about the unexpectedly high number of participants and in particular about the numerous experimentalists and local scientists attending the workshop.
The workshop program, as detailed below, comprised 14 invited communications (including 2 introductory lectures and 2 evening talks), 28 contributed talks, and a poster session with about 40 posters. The workshop also took the occasion to celebrate the 70th anniversary of Friedhelm Bechstedt and his seminal contributions to the field of theoretical spectroscopy.
The workshop started with a welcome by the local organizer, Claudia Rödl, pointing out the close relation between the city’s scientific past and presence and the workshop theme. As “city of light”, Jena was predestined to host this year’s ETSF workshop entitled “Light-Matter Interaction and Optical Spectroscopy from Infrared to X-Rays”. Building up on the historical achievements of Carl Zeiss and Ernst Abbe in optics and Döbereiner in materials science, Jena is nowadays a well-known science hub with strong research activities in these fields.
The opening session was dedicated to two introductory lectures. The first one given by Lucia Reining was a very didactive talk on “Fingerprints of correlation in electronic spectra of materials: How can calculations help to understand and make predictions?”. Reining structured her talk around the question: “Theoretical spectroscopy: how are we thinking?”. She detailed how first principles calculations contribute to our understanding, focusing on the main theoretical concepts, recent progress, and the comparison with experiment. Many materials properties are determined by electronic excitations. However, in matter an electron is never excited alone due to strong and long-term Coulomb interactions. From one-particle electron density approaches with approximated functionals, she directly went to Green’s-function-based methods to describe optical spectroscopy experiments. She explained that the GW method often fails for the description of satellite peaks, even though these should normally be included via the screened Coulomb potential. Reining discussed that the problem is not the lack of included interactions, but the way the electron-boson problem is solved: leading to the cumulant expansion. Since the latter is not an exact solution neither, she went even one step further by partitioning the system differently: in the T-matrix picture a different kind of dominating boson is chosen, going completely beyond the GW picture. Besides the scientific part, Reining added a very personal note by relating her talk to the life and merits of Friedhelm Bechstedt. She pointed out that Bechstedt has a very broad range of interests, resulting in multiple publications and three scientific books. He focussed on methodology, and screening was at the heart of his work. Very exemplary for young researchers, she emphasized that, what Friedhelm Bechstedt does, he does with enormous pleasure.
The second introductory lecture was given by Matthias Scheffler. He started with a very personal introduction of his friendship with Friedhelm Bechstedt, including a strong relation to German history during the cold war and the exceptional collaboration between East and West Germany through mutual visits of these two scientists. This was followed by the scientific part discussing the full ab-initio molecular dynamics description of thermal conductivity with a numerically efficient algorithm enabling convergence of the statistical mechanics in system size and time. He explained that thermal conductivity is a key characteristic of many materials, e.g. thermo-electrics, thermal-barrier coatings, or catalysts. However, it is largely unknown – of 225,000 identified inorganic semiconductor and insulator crystals, only 100 have any thermal conductivity data available. Besides this highly interesting science, he dedicated the last part of his talk to science policy. He presented a project to build a unique research-data infrastructure for sharing materials-science data (two proposals exist: NFDI and NOMAD). He insisted that research results and the related data should be open-access and that sharing will advance science significantly – provided that it is done correctly. The rules to do so have to be defined now. The proposed FAIR data infrastructure is based on four guiding principles: findability, accessibility, interoperability and reusability.
The first day of the workshop ended with a convivial welcome evening with drinks and some finger food in a local bar allowing the workshop participants to meet and socialize in a relaxed atmosphere.
The Tuesday morning session was dedicated to light-matter interaction and electron-phonon coupling. Emmanouil Kioupakis, in his invited talk, highlighted the important role played by lattice vibrations in mediating quantum processes of materials and offered an overview of the theoretical methods to describe phonon-assisted optical absorption in metals and in indirect-gap semiconductors such as silicon, boron arsenide, and hexagonal boron nitride (hBN). He also pointed out that the Auger recombination is an important nonradiative mechanism that dominates carrier recombination in wide-gap semiconductors. The results shed light on the interaction of light with materials and the efficiency of light-emitting devices; for that purpose, hBN was taken as an example. The peculiar optical properties of hBN have been addressed in detail by Lorenzo Sponza in the follow-up contributed talk. He reported on the excitonic dispersion in monolayer and bulk hBN, computed by solving the finite-momentum Bethe-Salpeter equation (BSE) both from first principles and with an excitonic model Hamiltonian. A strong momentum dependency of the exciton binding energy has been highlighted and pointed out as the origin of the asymmetry between absorption and emission spectra. The two speakers prepared the discussion on the optical properties in two-dimensional (2D) materials, animated by Friedhelm Bechstedt’s talk. He emphasized the fact that 2D materials are ideal platforms to explore new physical phenomena and reveal unique optical properties. By using the GW approximation and the BSE, he explored the response of graphene, graphene oxide, hBN, and molybdenum disulfide to a perturbing electromagnetic field and showed that it is dominated by in-plane light polarization. However, out-of-plane excitations in multi-layer systems are still allowed and enhanced despite the quantum confinement and the consequent reduced screening of the atomic sheets in normal direction.
The second part of the morning session dealt with the optical properties of transition-metal dichalcogenides (TMDs). In his invited talk, Bernhard Urbaszek opened the session providing a rich and exhaustive overview on spectroscopic techniques able to investigate the physics of excitons in 2D TMDs. The discussion on out-of-plane excitons was extended here with his preliminary results on ongoing experiments on TMDs monolayers. Then, he showed how the optical band gap and the photoluminescence emission time and yield can be tuned by changing the dielectric environment of the monolayers and using different substrate materials and heterostructure-fabrication techniques. Finally, the intra- and interlayer excitons in bilayers were examined highlighting the hole hybridization and magnetic-field dependence. The discussion on TMDs continued with Pedro Melo’s talk, with his special view on the impact of defects on the optical properties of tungsten disulfide monolayers, obtained by combining both GW and BSE. Such defects will inevitably alter the optical properties of the sample by changing its band structure (e.g. occurrence of mid-gap states) or the spin texture of the bands. Besides the defects, stacking atomically thin monolayers into multilayer or bulk systems affects the optoelectronic properties of TMDs. Philipp Marauhn drew our attention on this aspect investigating the spectral properties of TMDs by employing GW and BSE. A redshift of the optical absorption spectra with increasing number of layers is theoretically observed in agreement with experiments, due to an enhanced screening environment effect. In addition, as a consequence of stacking, intralayer or interlayer excitons occur in mono- and multilayer or bulk systems, respectively.
After lunch, the workshop turned page to electronic-structure methods, with a special focus on electronic correlation. Pina Romaniello, with her invited lecture on photoemission processes, provided us with a panoramic view on the theoretical methods to get insight into the electronic structure and excitations in correlated materials. Although many-body perturbation theory within the GW approximation is the standard method to calculate photoemission spectra of many materials, it does not capture strong correlation, unless one treats the system in a magnetically ordered phase. In the talk, she showed that the knowledge of the lowest n-body reduced density matrices (RDMs) can provide accurate photoemission spectra in model systems in the weak as well as strong-correlation regime. An application to several transition-metal oxides was given. Afterwards, Savio Laricchia drew our attention to the fact that density-functional theory (DFT) underestimates electron-phonon interaction and that, on the other hand, nonlocal approaches to exchange and correlation (hybrid functionals and quasiparticle GW) enhance it. He introduced a new methodology to treat electron-phonon interaction. The approach, implemented within the quasiparticle self-consistent GW formalism, succeeds in describing the electronic properties for a wide range of materials, including many where standard DFT fails.
Motivated by the evidence that newly developed functionals in DFT improve the energy estimation but worsen the error in the density, Martin Panholzer, suggested in his talk to improve the accuracy of ab initio methods by starting from a Jastrow-correlated wave function, i.e. a product of the correlation factor and a Slater-determinant. First, the correlation factor is optimized by employing cluster-expansion techniques, followed by the optimization of Slater-determinant orbitals. The real challenge is to implement the coupled optimization of the correlation factor and the Slater-determinant for 3D materials as an extension of present DFT implementations. Preliminary results on 1D model systems were shown. The discussion on correlation was continued by Charles Patterson. He reported on an all-electron Gaussian-orbital Coulomb-metric density-fitting method to accelerate SCF and post-SCF calculations, including GW and BSE. He analyzed the performance of the density-fitting method with respect to an auxiliary fitting basis. The method was illustrated by an application of time-dependent Hartree-Fock and GW/BSE calculations on Na clusters with up to 20 atoms and on simple covalent solids using the Exciton code. The afternoon session was closed by Nicole Helbig. She presented her study about the relation between correlation and localization for hydrogen chains of different lengths and compare the reduced density-matrix functional theory (RDMFT) results to those from Monte-Carlo simulations.
Right after, the poster session with about 40 posters covering all the topics of the workshop started, accompanied by drinks and a Thuringian barbecue that was served in the courtyard of the Rosensäle building. In a relaxed atmosphere, the participants took the occasion for informal and yet intense scientific discussions around the posters. The poster session lasted 4 hours in total, significantly longer than originally foreseen in the program. The posters were displayed during the entire duration of the workshop.
The morning session, dedicated to nonlinear optics, was opened by Myrta Grüning’s invited lecture. She reported on real-time approaches derived from a nonequilibrium Green’s function formalism that are capable of describing light-matter interaction beyond the linear regime, fully accounting for excitonic effects. The proper inclusion of these many-body interactions has been highlighted for the case of single-layer monochalcogenides, whose strong second harmonic generation cannot be reproduced within the independent-particle approximation. Valérie Véniard presented her work on the ab initio calculation of the second-order susceptibility for describing the linear electro-optic (LEO or Pockels) effect, i.e. the change of the refractive index proportional to an applied DC electric field. Veniard’s talk covered the electronic and ionic contributions to the LEO tensor, which were evaluated within the framework of TDDFT with approximate treatments of quasiparticle and excitonic interactions. The screening of electronic interactions was a central subject of the talk by Thorsten Deilmann, focused on the practically relevant case of nanostructures (e.g. molecules, clusters) on metallic substrates, where the latter strongly affect the optical properties of the adsorbate. Deilmann showed that in approaches where the states of the substrate only contribute to screening, but are not explicitly included in the BSE Hamiltonian, it is important to screen both Coulomb and exchange interaction. This latter element turned out to be essential in order to obtain the correct quasiparticle energy ordering and optical gap.
The quest for an improved description of exchange-correlation interactions in the framework of (TD-)DFT has been the subject of the talks by Ayoub Aouina and Rex Godby. In both talks, novel nonlocal and nonadiabatic exchange-correlation potentials or kernels were introduced, either by defining systematic routes for mapping model results onto a real system, or by comparing the exact response function for various interacting systems and its counterpart for the auxiliary system of noninteracting electrons. The morning was closed by the ESTF general meeting. The meeting introduced the concept of ETSF to those attendees that are not ETSF members and provided an update on the network activities, such as the e-learning platform, collaboration teams, and the new website.
The afternoon sessions were dedicated to applications of electronic-structure theories to optoelectronics and energy generation. The invited talk by the experimentalist Koen Vandeval on organic photovoltaics set the stage for the subsequent discussion. Vandewal emphasized the fact that, in spite of the low dielectric constants, electrons and holes split efficiently in organic solar cells. However, large amounts of energy are lost in this process resulting in low voltages that can be extracted from the cell. By reviewing a large body of experimental data, and in particular by analyzing light-absorbing and emitting charge-transfer states that mediate the charge separation, the detrimental effect of nonradiative energy losses arising from molecular high-frequency vibrations has been identified, suggesting strategies for the optimization of device performance. The theme of intra- and inter-molecular molecular vibrations in organic systems of interest for optoelectronic applications has been further developed by Frank Ortmann. In particular, his talk focused on the effect of molecular flexibility and polaron formation on the excitonic properties and their temperature dependence. Ab initio and model Hamiltonian approaches were proposed to describe the above mentioned phenomena. The discussion then moved to hybrid lead-halide perovskites for photovoltaics and light-emitting applications with the talk by Claudio Quarti, focused on the puzzling signatures of exciton-phonon coupling in resonance impulsive stimulated Raman spectra as unraveled by ab initio lattice dynamics calculations.
The talk by Roberta Poloni discussed photo-switchable molecules in the perspective of their application in photoactive metal-organic frameworks (MOFs) for efficient gas separation. Poloni first focused on the optical properties of azobenzene molecules in solution, for which embedded GW/BSE calculations provide an excellent agreement with experimental data. A similar fragment approach was then applied to azobenzene-functionalized MOFs, investigating the photophysics in a confined anisotropic environment. The comparison to periodic GW/BSE calculations revealed that the fragment approach is suitable for describing localized excitations on the functionalized ligands of the MOF. The last talk of the session was given by an experimentalist, Marco Grünewald, who spoke about optical in situ spectroscopy of ultrathin organic films by combining differential reflectance spectroscopy and photoluminescence, thus allowing for quasi simultaneous measurements of optical absorption and emission properties. He exemplified his method for PTCDA on mica showing that substrate treatment has a strong impact on the structural and electronic properties of the adsorbate.
Wednesday finished with two evening talks (or better late-afternoon talks) given by local scientists from the Friedrich-Schiller-Universität Jena who had been asked to introduce their fields of research in an entertaining way to nonspecialists. Gerhard Paulus spoke about a newly developed experimental technique for non-invasive cross-sectional imaging of semiconductors by transferring the principle of optical coherence tomography to the x-ray regime. He demonstrated that this technique is sensitive to buried layers of only a few nanometers of thickness and opens great potential for device characterization. Christian Eggeling continued with an overview on the physical principles behind superresolution microscopy showcasing its merits and potential, especially for biomedical applications. In particular, he focused on long-standing questions regarding the structure and functioning of lipid membranes in biological cells. Both speakers were extremely pedagogical and managed to bridge the gap between the ab initio community and their respective fields of research.
The morning of Thursday was dedicated to ultrafast phenomena, especially related to photoemission and excitons. The invited presentation by Luca Perfetti opened the session by presenting his pump-probe experiments, with a specific focus on 2D electron gases obtained by depositing Cs on InSe. Femtosecond-resolved experiments showed that the deposition of Cs triggers a transition to a 2D electron gas, switching from classical to quantum behavior, the latter being characterized by hot out-of-equilibrium carriers. Next, Eugene Krasovskii brought us down to the attosecond timescale, with his results for electronic band structure aspects in the interaction of outgoing electrons with the crystal lattice. Felix Otto reported on the capabilities of photoemission-momentum-map experiments to access molecular orbitals, reporting on the relevant case study of polycyclic aromatic hydrocarbons on noble-metal surfaces.
After the coffee break, Davide Sangalli’s invited talk introduced the audience to recent theoretical developments in exciton dynamics, in tight connection with pump-probe experiments. Making use of a Green’s function formalism, exciton dynamics and exciton lifetimes were studied. Antonios Alvertis continued with this topic studying the effects of exciton-phonon coupling and thermal expansion to comprehensively understand the temperature dependence of exciton energies in molecular crystals. Then, Hsiao-Yi Chen focused on exciton dynamics and radiative decay rates, showing that ab initio results are nowadays able to determine radiative lifetimes within a factor of 2 from experiments.
Stefanie Gräfe opened the afternoon session on electronic-structure methods dedicated to optics with an invited talk on strong-field physics and chemistry. She presented numerical solutions of the time-dependent Schrödinger equation outlining various handles for controlling electron dynamics in molecules by applying an external laser field. Her presentation was followed by the two contributed talks by Jaakko Koskelo and Christine Giorgetti, the former presenting latest improvements and shortcomings of the BSE treatment for the homogeneous electron gas, and the latter raising an intense debate between the participants about the best methodology to treat finite systems (here, slabs) within a supercell approach.
The talks by Christoph Friedrich and Beata Taudul closed the afternoon session. Friedrich brought us to magnetic phenomena, discussing the treatment of magnetic excitations by the BSE and how the use of incomplete summation in the self-energy could lead to broken causality. Taudul’s talk was about the use of core-level excitations, hence bridging with the Friday morning session, to determine through a joint experimental-theoretical effort the order of Li removal from crystal structures in Li ion batteries.
After the scientific program, the workshop participants had the possibility to join a guided tour through the city center of Jena that put the town’s scientific history into the greater context of general history. The day ended with the social dinner that comprised the official birthday party for Friedhelm Bechstedt.
The last day of the workshop was dedicated to core-level and x-ray spectroscopy. Claudia Schnohr opened the day with an invited talk on x-ray absorption spectroscopy of semiconductors alloys. The latter are widely used for electronic, optoelectronic, and photovoltaic devices because their materials properties can be specifically tailored by adjusting the material composition. However, alloying often introduces a discrepancy between element-specific short-range properties and average long-range properties. She demonstrated how x-ray absorption spectroscopy can be applied to study element-specific local structural and electronic properties of semiconductor alloys as a function of composition. The subsequent talk by Hazem Aldahhak focused on the oxidation state of manganese corroles on Ag(111). This class of molecules is interesting for catalytic applications, stabilizing high-valent metal states. Despite being crucial for their functionality, their electronic structure is often hardly accessible with traditional techniques, specially under interfacial conditions. He presented a multi-technique strategy based on DFT, element-specific x-ray spectroscopy, and in-situ atomic force microscopy to unravel the electronic structures and the oxidation states of prototypical corrole complexes as well as of their thermally induced derivatives adsorbed on Ag(111). Eventually, Vasil Rokaj explained that much theoretical work has been devoted to describe properties of periodic systems with arbitrary homogeneous magnetic fields in different regimes, however, a general first-principles modeling of such fundamental effects is lacking. He proposed a solution to the problem of Bloch electrons in a homogeneous magnetic field by including the quantum fluctuations of the photon field and presented a generalized quantum electrodynamical (QED) Bloch theory from first principles.
All-electron many-body perturbation theory for core spectroscopy in organic, inorganic, and hybrid materials was the subject of the invited talk by Caterina Cocchi which resumed the session after the break. She focused on all-electron schemes, offering the additional opportunity to calculate core excitations on the same footing as optical ones. After introducing the underlying methodology, she demonstrated the predictive power of this approach for computing x-ray absorption spectra and excitations with a few recent examples on different material classes, ranging from bulk crystals to organic semiconductors and hybrid inorganic-organic perovskites. Next, Igor Reshetnyak presented ab-initio optical data and inelastic x-ray scattering spectra of liquid water calculated within the frameworks of BSE and TDDFT. From the extensive comparison of theoretical results with available experimental data, information concerning the electronic screening has been extracted, allowing to reduce the ambiguities in the liquid water theoretical band-gap value. The last presentation was held by Christian Vorwerk on electron-hole correlation in resonant inelastic x-ray scattering (RIXS) of solids. In his talk, he presented a novel many-body approach to RIXS, making use of excited states in both the optical and x-ray region, obtained from full diagonalization of the BSE. An all-electron framework has been employed to describe core and valence excitations consistently. He showed a reformulation of the Kramers-Heisenberg formula in terms of pathways between excited core and valence states in a many-body picture and how these excitation pathways allow for an intuitive interpretation of the electron-hole correlation in the RIXS process.
The closing remarks of the workshop were given by Silvana Botti, pointing out the high scientific quality of contributions and discussions. She highlighted the large number of participants and mentioned that also many non-members of the ETSF network were present, as well as participants from non-EU countries. At the very end, Gabriele D’Avino announced the ETSF workshop 2021 which will be held in Leuven (Belgium).